Digging for cinematic channels
Many technologies now call for deeper manipulations of video streams. The recent HTML5 welcomes video as an integral part of the web through the <video> and <canvas> tags. Javascript libraries are stimulating the imagination for play with motion graphics. Along these developments, the field of computer vision continues to astonish with what is possible to extract from flowing pixels. It is still challenging to see how open web standards will merge with raw video processing but the potential can already be perceived. Most particularly, in the mobile context, Augmented Reality promises to compose extra feeds upon the audiovisual captures of portable devices.
Considering the cultural implications of such practice, I like the idea that the reverse is also possible. I like the idea of techniques to balance out and open spaces for creativity through a constant stream of visual data. Those techniques come from a branch of computer vision that proposes to reduce the noise and “diminish” reality. Labs approach the idea using different methods. I couldn’t tell which offers the best results between the methods proposed by the University of Florida or the Technical University of Ilmenau in Germany. But I realise they both call it “Diminished Reality“: The process of removing objects from video.
Can practical applications of “open video” lie within such development ? The possibility to open video spaces. Dig for cinematic channels. Regenerate canvas for new compositions. «Removing objects from video» to create ground for new connexions.